This study investigates the role of encrypted messaging apps, specifically WhatsApp, in disseminating political disinformation during South Africa’s 2024 general elections. Drawing on a dataset of 22,384 messages from 47 politically active WhatsApp groups and 6,283 unique users, the study identifies how AI-generated deepfakes, emotional manipulation, and ideological group structures facilitated the spread of false political narratives. Using MyFactChecker, an AI-powered sentiment and verification tool, we reveal that disinformation most frequently relied on fear-based appeals (41%), identity-driven rhetoric (32%), and content mimicking credible journalism (27%). The findings show that disinformation gained traction not through factual accuracy but through emotional resonance and relational trust within ideologically cohesive groups. Even when flagged as false, corrective content was often dismissed as propaganda or foreign interference, thus underscoring how truth contests unfold within closed, affective networks.
WISER Seminar Papers
- WISER's TRUST seminar is hosted on-line every Monday afternoon from 16:00 - 17:00 SA time during the teaching semester. Forthcoming seminars are available here, and past events are detailed in our archive.
- Please register on Zoom in advance of the meeting on this link.
- Participants should please read the paper (below) prior to the seminar; it will typically be available by the Friday preceding the seminar.
Presented by : Gregory Gondwe
18 May 2026 - 4:00pm
Paper for discussion:
WISER Research Theme:
